An A/B test compares two newsletter versions that differ in one element. Assign recipients to groups at random where possible and decide before sending which metric answers your question.
How do you plan a fair comparison?
If you test two subject lines, keep the sender, content, and send time as similar as possible. The groups should not overlap. Choose a metric that matches your goal. Recorded opens have limitations because email clients can load images automatically. If your goal is visits to an offer page, click rate may be more useful.
Example with no clear winner
A team sends two variants at about the same time to randomly assigned groups, each with 100 delivered emails. It chooses click rate in advance. Five recipients click in version A and six in version B, giving five and six percent. At these group sizes, one additional click is not enough to establish a winner.
How does mailfino split an audience for an A/B test?
In mailfino, you can split a list into two groups. Prepare the versions separately, send them separately, and compare the metric you selected in advance.
